Monday night the Oakland A’s gained a playoff berth with a roster that most experts believed at the beginning of the season would not even approach playing .500 baseball. Although most experts in hindsight now see the quality of the A’s players, they are failing to realize their success is again the result of Billy Beane’s (Oakland’s General Manager) ability to use analytics to gain a competitive advantage.
In the book (and subsequent movie) Moneyball, which was published in 2003, Michael Lewis showed how Beane used analytics to assemble a competitive baseball team, despite having less revenue than most of the clubs he was competing against. Beane understood that traditional statistics used to judge players and prospects was flawed and there were better predictors of success.
The A’s most recent success, highlighted by reaching the playoffs, reinforces the advantage analytics can bring. Again, Beane assembled a team of largely unknown and unheralded players. It is understanding that many “experts” attribute the A’s success to Beane abandoning Moneyball because he is no longer focusing on the traits described in the 2003 book. Although the metrics where these players excel in are different than those publicized in 2003, given the lack of star power (very few of his players were considered key by other clubs) and the fact most of the players share common skills, it is clear that he again found a formula that allows the A’s to score more runs than their opponents. Although I have no idea what metrics Beane used to optimize performance, he obviously found an algorithm that allowed him to collect players that fit his budget and were still able to generate a playoff berth.
The other interesting takeaway is you do not see Beane talking about using analytics or leaking stories about it to the press. After Moneyball, many major league baseball teams (most with more financial resources) copied his approach. When everyone is using the same analytics to judge players, it is no longer a competitive advantage. That is one of the main reasons that for several years the A’s have not been able to field a playoff-caliber team in recent years. My guess is Beane will not be inviting Michael Lewis into the A’s clubhouse to write a new book about their new approach.
There is a clear connection between how Beane approaches assembling a baseball team and how a social game company uses analytics to build a game. The principle behind Beane’s approach is that he needs to put together a team that would regularly score more runs than its opponent. He would do this by reverse engineering how a team scores runs (gets on base, advances bases, etc.) and then finding players (within his budget) who will generate more run production than they allow. This could be achieved by scoring more or limiting their opponent’s run production. This is similar to the equation that social game developers and publishers face; they need to generate a higher LTV (lifetime value) than their cost of user acquisition (CPI or CPA, depending on the formulas they are using).
There are some key takeaways for social game companies from the A’s success:
- Analytics can help you win even if you have limited resources. Using analytics effectively can allow you to compete with companies that may have two or even ten times the cash.
- Constantly reinvent. The optimal equation a year ago or now may not maximize your revenue tomorrow or next month. You need to reevaluate constantly your inputs, equations and how you look at the opportunity.
- It is not a competitive advantage when everyone does it. If you are doing what every other social company already does as part of its business, it is part of the price of doing business, not a competitive advantage. You need to find unique or better ways to use analytics than your competitors.